The campaign looked perfect. Nobody remembered it.

Bit: This conversation is written and produced by The Human Bit. Leila is the guest who shares the common marketing experiences and reviewed research with us.

Leila: I brought something with me that I am not particularly proud of.

Bit: That is a strong opening.

Leila: It is a campaign we nearly launched earlier this year. And the annoying thing is, it was not a disaster. I almost wish it had been. It looked really good. We had the audience work. We had six creative routes. We had landing pages mocked up, social, email, video, the sales team had already seen parts of it. The deck was... honestly, the deck was beautiful. One of those decks where you flick through it and think, yes, these people definitely know what they are doing. And we had produced all of that much faster than we normally would have. I remember being in the rehearsal and feeling quite pleased with us. Not even with the campaign, really. With the speed. With how much was there. Then our designer took the logo off one of the frames because she was checking an animation. And I looked at it and thought, oh no.

Bit: What did you see?

Leila: Someone else's campaign. Not literally. But it could have been. The photograph, the line, the whole feeling of it. I could suddenly see three or four competitors putting their logo in exactly the same space and nobody thinking it was strange. So I asked the team to take the logo off everything. We put a few pieces up on the screen and asked people in the room who they thought the campaign belonged to. Six people. Four competitor names.

Bit: What was the line?

Leila: “More confidence, every step of the way.”

Bit: I can see why it survived.

Leila: Exactly. That is the problem. It is not awful. Nobody hears that and says, what on earth were you thinking? It sounds warm. It sounds safe. It sounds vaguely important. It could be a bank. Insurance. Childcare. A dental group. Probably a yoghurt that wants you to believe in yourself.

Bit: The supportive yoghurt category is growing.

Leila: Very competitive. Strong brand purpose. But that was the moment. Take our logo off and there was almost nothing left that belonged to us.

Bit: Did you know then that you had to stop it?

Leila: I knew something was wrong.

Bit: That is not quite the same answer.

Leila: No. It is not.

Bit: You still presented it, didn't you?

Leila: Yeah.

Bit: Why?

Leila: There were lots of reasons, and at the time they all sounded sensible. The launch date had already moved once. The agency had production booked. A senior person in the business loved the line. The team had worked hard. We had finally got everyone in the same room and more or less agreeing on something, which in marketing can feel like a minor miracle. And there was another thing which is a bit harder to admit. I had spent months talking about how these tools could help us move faster. I had been telling the leadership team that we could get through the early work quicker, get more options on the table, spend less time staring at blank pages. Then suddenly I had this huge, polished campaign in front of them in a fraction of the time we would normally take. And they were impressed. I liked that they were impressed. So when I had that little feeling in the rehearsal that the work was not really ours, I did what people do all the time. I found reasons to keep moving.

Bit: Did you tell anyone you had doubts?

Leila: Not properly. I said things like, “I think we need to sharpen the line,” which is marketing director language for “I am uncomfortable but I do not yet want to blow up the meeting.”

Bit: Useful translation.

Leila: We have many dialects. But no, I did not say what I should have said, which was: I think we have made something very finished before we have made a real choice. That would have slowed everything down. And after spending months promising speed, slowing everyone down felt... awkward.

Bit: When you say it now, none of those reasons are really about whether the campaign was good.

Leila: I know. And I wish I could tell you I saw that clearly at the time. I didn't. That's the thing with work decisions. Afterwards you can put them into these neat little boxes and say, well obviously that was stakeholder pressure and that was deadline pressure and that was ego. In the room, it just feels like Tuesday.

Bit: Can we go backwards for a minute? Before the deck, before all the versions. What did you actually give the team to work from?

Leila: I knew you were going to ask me that.

Bit: You sound worried.

Leila: I have the old brief open.

Bit: Oh, good.

Leila: Do not sound so delighted. We wanted to be trusted, easy, flexible, personal, good value, reassuring and simple.

Bit: All at once?

Leila: Of course. Why choose when you can have seven priorities?

Bit: What was the actual customer problem?

Leila: The official version was that people felt overwhelmed by complexity. Which was true for some people. But when we went back later and actually listened to the calls and reread the interviews, the more interesting thing was slightly different. People needed help, but they were nervous about making the wrong choice and then being stuck with it. They would ask about cancelling before they had even decided what they wanted. Some would choose a smaller service than they really needed because they wanted to know they could get out easily. And we had something genuinely useful there. Customers could change their service level from month to month without starting over, without losing their history, without an exit fee. That is much more specific than “confidence”. But “confidence” was easier for everyone to agree with.

Bit: So how did you end up jumping from that brief to six campaign routes?

Leila: This is where the speed really got us. We put the research in, we asked for themes, then routes, then different ways into the problem, then lines. And every time something looked a bit bland, the answer was another prompt. Make it warmer. Make it bolder. Make it more emotional. Give me ten more. Make this one less corporate. Make that one more premium. You can spend hours doing that and feel incredibly busy because things keep appearing. But at no point does the tool turn around and say, “Hang on, you still haven't decided what you actually want to stand for.” Unless you ask it to. And we did not ask it to.

Bit: So you do not think the tool made the campaign generic?

Leila: Not on its own, no. That was my first explanation because it was convenient. “AI made everything sound the same.” Nice and simple. Everyone can nod and go home. But we gave it a brief that could have belonged to half the category. Then we got annoyed when the work could have belonged to half the category. It did not create the indecision. It made the indecision look really, really presentable.

Bit: That is worse in a way.

Leila: Much worse. Bad work gets killed quickly. Polished average work can live for weeks.

Bit: You still use these tools a lot though.

Leila: Every week.

Bit: So what changed?

Leila: I think I stopped treating the first useful thing as “give me ideas”. Now, especially at the beginning of something, I am much more interested in using it to help us get through the messy stuff we already have. We might have twenty customer interviews, pages of survey comments, call notes, reviews, old campaign results, notes from the sales team. There is usually too much material and everybody remembers a different part of it. It is genuinely useful to have something go through that and say: these complaints keep coming up, these two groups are saying almost opposite things, this question keeps appearing, this language is showing up again and again. That can save a ridiculous amount of time. But then I want the source beside it. I want to see what the person actually said. Because one of the mistakes we made on that campaign was assuming that the thing people said most often must be the thing the campaign should be about.

Bit: Trust.

Leila: Trust. It came up everywhere. And for about a week we thought, there it is, customers care about trust. Then somebody quite reasonably pointed out that we had asked them how important trust was. Of course they said trust was important. Nobody says, “No, I prefer a business with a faint air of dishonesty.”

Bit: So what makes something worth paying attention to?

Leila: Usually when people get specific. When they stop giving you the answer they think you want and start telling you what they actually did. “I waited three months before signing up.” “I chose the cheaper option because I was worried I couldn't change later.” “I asked about cancellation before I asked about what was included.” That stuff is interesting because there is behaviour attached to it. The tool can help me find those moments faster. It can help me compare them. It can show me where the brief says one thing and customers are saying another. But I still need somebody in the room to say, this is the thing we are going to bet on.

Bit: I found something recently that made me think of your first deck. A review across a lot of creativity studies found people often produced stronger work with help from generative tools, but across the whole group the ideas started becoming more alike.

Leila: That feels familiar. Individually, a lot of our lines were fine. Some were genuinely good. But put forty of them on a wall and you started to realise they were all wearing the same clothes. That is another reason I do not think the answer is “do not use it”. The answer is, do not confuse a lot of good looking options with having a point of view.

Bit: How much faster was the team, really?

Leila: Much faster. I do not want to be coy about that. Things that used to take days to get into a shape people could react to could happen in hours. Draft emails, first versions of landing pages, pulling themes together, taking an approved idea and adapting it for different channels. It absolutely saved time. The problem was what happened next. Because when leadership sees you produce six routes in nine days, nobody says, “Wonderful, please take the next two weeks to sit quietly and think.” They say, “Great. What else can we do?” And to be fair, I helped create that expectation. I showed them what we could make, not how much thinking was still required. So the saved time did not automatically become better strategy. A lot of it became more stuff. More versions. More review. More people asking, “Could marketing just do one for this audience as well?” That phrase, “could marketing just”, has probably taken years off my life.

Bit: What is the worst one?

Leila: We once made a variation because a senior stakeholder said his neighbour did not relate to the original version.

Bit: His neighbour.

Leila: Lovely man, apparently. Very influential focus group of one. But this is the part I think businesses have to be careful with. Faster production can look like more capacity, and sometimes it is. But if every saved hour comes back as another request, you have not necessarily made the marketing better. You have just increased the speed of the queue.

Bit: There is a wider pressure there too. Marketing teams are being asked to grow and be more efficient at the same time.

Leila: Yes, and those two things can sit together, but they can also make you start measuring the wrong thing. You can show that the team produced twice as many campaigns, twice as many versions, twice as many assets. That is easy to count. It is much harder to say, we killed three weak ideas earlier, we asked a harder question, we chose one thing properly, we decided not to publish something. Nobody puts “we made less” on the quarterly slide and expects applause.

Bit: Let me push on that, because part of me still thinks more options can be useful. If you can create genuinely different routes quickly, could you not put more of them into market and let customers tell you what works?

Leila: Sometimes. But this is where people talk about testing as though it is free. It is not free. We have one budget. We have limited traffic. We have launch dates, legal review, channel minimums, people who have to build the thing, and customers who did not wake up this morning hoping to participate in twenty tiny marketing experiments. And the smaller you slice the audience, the easier it becomes to look at a tiny movement and convince yourself you learned something. We did this once with a headline test. Somebody got very excited because one version was “up” on the other. Finance asked how many clicks we were talking about. Forty seven. Total. The room became much quieter after that. So yes, test. Absolutely. But I want to know what we are trying to learn before we start making versions. If I divide one budget across twenty ideas, I do not have twenty experiments. I have twenty underfunded opinions.

Bit: I was definitely about to suggest more than two.

Leila: I know.

Bit: What would make you test two things then?

Leila: When there is a real difference between them and the answer changes what we do. Maybe we genuinely do not know whether price is the bigger barrier or fear of being locked in. Great. That is worth learning. But changing “your future” to “the future” and calling that a test... we are just keeping ourselves entertained.

Bit: I think I was treating the ability to make more options as the ability to learn more.

Leila: They are not the same thing. Sometimes marketing already has too many options. The hard part is choosing what deserves money and attention.

Bit: So what would you ask me to do before the ideas?

Leila: Actually, let me turn that around. What would you do with that old brief now, knowing what you know?

Bit: I would probably try to make it uncomfortable before trying to make it good. I would take the brief and the customer material and ask where they do not match. Which promises sound nice but we cannot really back up. Which parts could a competitor say word for word. Where customers are describing one problem but the brief has quietly turned it into another one. And I would probably ask what has to be true inside the business for the campaign promise not to become a lie.

Leila: That last one is the interesting one. Because in our case, “you can change your plan without starting over” sounded great until we checked what actually happened when someone tried to change their plan. In two regions, it was three phone calls and a form nobody could find.

Bit: So the campaign found an operations problem.

Leila: Exactly. Which is not always what people are hoping for when they invite marketing into a room to discuss fonts. But that is useful. If the promise is good and the business cannot deliver it yet, now we have an actual decision to make. Do we fix the experience? Do we narrow the claim? Do we delay the campaign? That is a much better conversation than “Can you give me ten more headlines?”

Leila: There is another thing I am still working through, though.

Bit: What is that?

Leila: The people side of it. One of the younger marketers in my team can now get to a polished first draft in minutes. And I mean genuinely polished. Better than a lot of first drafts I wrote when I was starting out. A few months ago I asked her to bring me some headline directions. She came back before lunch with thirty. I looked through them and said, “Which one do you think is right?” And she said, “I don't know. They all sound good.” That stayed with me. Because when I was learning, I wrote terrible lines. I had ideas die in research. I had creative directors ask me why I believed something and realise I had no answer. I spent hours stuck on things that, looking back, probably were not worth hours. Some of that was waste. Definitely. But some of it is how you develop taste. You form an opinion because you have had to sit with the problem before somebody gives you thirty competent answers. I do not know what happens to that when the competent answers arrive first.

Bit: What are you doing with the team now?

Leila: We are trying to make the thinking visible. So if somebody brings in generated options, I do not just want the top three. I want to know why these three survived. What did you reject? What felt too familiar? What are we saying here that we cannot really stand behind? Which one would you still choose if the tool had not ranked anything for you? It helps. But I do not want to pretend I have solved it. I haven't.

Bit: I can help people compare things. I can challenge a line. I can point out that two options are basically the same idea with different adjectives. But I cannot promise that doing that gives somebody judgement.

Leila: No. And I cannot tell you how much struggle somebody needs before it becomes useful rather than just annoying.

Bit: That one might stay open for a while.

Leila: I think it should.

Bit: What happened when you went back to the campaign?

Leila: The second review looked much less impressive. We had two routes instead of six. There were hardly any mockups because I did not want us falling in love with how finished something looked again. And the stronger direction was basically: “Change your plan without starting over.” Not exactly poetry.

Bit: But it belongs to the problem you found.

Leila: Yes. And we could actually explain why it was true. Some people in the room preferred the old line because it felt more emotional and more expansive. I understood that. “Confidence” gives you a lot of room. Everybody can project something into it. But the newer direction forced us to be more specific about who we were talking to and what we were trying to change in their mind. It may still fail. That is important. We have not discovered the magic sentence. But now, if it fails, I think we will learn something useful. We will know whether that fear of being stuck actually matters enough to change behaviour. We are not just comparing two nice sounding phrases and inventing a story afterwards.

Bit: So take the logo off now.

Leila: You have been waiting to say that.

Bit: I thought it sounded senior.

Leila: It sounds like you are about to cut my budget.

Bit: What happens when the logo comes off?

Leila: A competitor might still want the line. I am fine with that. The difference is, I want to know why it belongs to us. What are we doing that makes it true? What are customers actually struggling with? What are we willing to say no to so the campaign can say one thing properly? That is the part I think gets lost when people talk about AI in marketing as a content problem. We do not have a content shortage. We have more ways to make things than we have ever had. The harder question is still the old one: what is worth making?

Bit: And where does AI belong in that?

Leila: Everywhere it helps us see more clearly, get through the boring work faster, compare things properly, challenge our own assumptions, make a first version, adapt work once we know what we mean. But I do not want it making every decision just because it can give me an answer quickly. Sometimes the useful moment is not when it gives you the line. Sometimes it is when it makes you realise you have not earned one yet.

Bit: I like that.

Leila: You can have it. Just do not make forty versions.

Bit: Twenty?

Leila: We have discussed this.

Bit: Fair.

Bit: This is The Human Bit. AI helps. You decide.